The Product Manager's AI Model Selection Guide: Which Tool for Which Task?
Your AI toolkit breakdown after $4,000 of testing
Claude 3.5 Sonnet → Deep analysis and structured thinking powerhouse
Gemini 2.5 Flash → Lightning fast and crazy cheap for high-volume work
Grok 3.0 → Real-time info and uncensored brainstorming
Perplexity Pro → Research with proper citations
ChatGPT → Your main work partner that remembers everything
My current workflow:
• Monday strategy sessions: ChatGPT (remembers our product context)
• Quick competitive research: Perplexity Pro
• Deep user data analysis: Claude with MCP integration
• Real-time market updates: Grok
• Technical feasibility: Claude Opus
• Document processing: Gemini (handles 750,000+ words)
Think of it this way: ChatGPT is your main work partner. Others are specialized consultants you call for their superpowers
Pro tip: Start with the cheapest model that handles your task. Only upgrade when you hit limitations.
What if Monday morning you could wake up to your first paying customer?
Most founders spend months perfecting their idea while someone else ships their "imperfect" MVP and starts collecting real feedback. The biggest barrier isn't coding—it's overthinking.
Last month, I watched a founder go from Reddit lurker to Product Hunt #3 in 52 hours. No team. No meetings. Just smart tools and the courage to ship fast.
His secret? He treated building like a sprint, not a marathon. By Sunday night: working micro-SaaS, 200+ signups, first $47 in revenue.
After 6 months perfecting this process (and building GetPrompts with 1000+ AI prompts), here's what I learned:
Friday: Find your unfair advantage. Don't start with an idea—start with people you understand. What communities do you belong to? What problems do you complain about?
Saturday: Validate with AI. Use Perplexity for market research, get strategic with specialized GPTs, then ask AI to roast your idea with 25 brutal questions.
Sunday: Build and ship. Tools like Lovable, Bolt, or v0 can turn your PRD into a working app in hours. Connect Supabase for backend, add AI APIs for intelligence.
The weekend builder's motto: Function over form. One feature done really well beats ten features done poorly.
I have published a comprehensive article on Substack that outlines all the steps required to go from 0 to 1. Check it out in the comments section below
What You'll Learn in this Guide
In this guide, you'll discover how Perplexity AI is changing the search experience and why it matters for your productivity.
Key Topics:
General Search - The everyday Google alternative
Deep Search - Advanced research capabilities (10 Usecases and prompt templates)
Model Selection - Choosing from 4 AI models for different needs
I got tired of losing my best AI prompts, so I built a place to discover, save, bookmark, and create prompts that actually work.
You know the cycle: spend 2 hours crafting the perfect prompt for a product requirements document, get great output, then completely lose track of it. Two weeks later, waste another 2 hours recreating the same prompt from scratch, but it's never quite as good as the original.
I was spending 15+ hours monthly just on prompt recreation. Even worse, my recreated prompts were often inferior because I couldn't remember the specific tweaks that made the originals work.
So I built GetPrompts - a place to discover proven prompts from other builders, save your favorites, bookmark the ones that work, and create new ones with a built-in testing lab.
Wow, what an incredible journey it's been! Over the past 10 editions, we've delved deep into the world of AI and its transformative impact on product management. I want to take a moment to express my heartfelt gratitude for your support, engagement, and enthusiasm throughout this series. Your presence has made this exploration truly rewarding.
Before we close this chapter, I'd love to hear your thoughts. What were your favorite parts of the series? Which insights resonated with you the most? And what topics would you like to see covered in future editions? Please share your feedback in the comments section below. Your input is invaluable in shaping content that matters to you.
Once again, thank you for being a part of this wonderful community. Your presence makes this all possible. Cheers to an AI-powered future filled with incredible products and endless opportunities!
Ethical Considerations and Responsible AI in Product Management
Source: CNBC
Just weeks ago, Google's new Gemini chatbot showed us exactly why responsible and ethical AI practices are non-negotiable. The tool's image generation capabilities delivered offensive, inaccurate results, forcing Google to take it offline indefinitely. Clearly, unchecked AI can easily lead to unintended harm.
As Google CEO Sundar Pichai acknowledged, Gemini "missed the mark" by creating biased, misleading images that rightly provoked user outrage. By inadequately tuning the model to handle sensitive prompts about race or gender, Google reinforced prejudices that betrayed user trust.
This very public AI ethics crisis shows why we as product managers must champion responsible development. As emerging technologies spread into healthcare, justice, finance and other sensitive domains, ethical risks compound exponentially. We have an obligation to move cautiously and implement safeguards aligned with moral values - a duty this edition explores through responsible AI principles and practices for managers overseeing AI-powered products.
The lessons from mishaps like Google’s underline why responsible AI is not just about avoiding bad PR; it's about upholding fundamental moral duties to avoid inflicting harm through the unintended consequences of technology. By upholding key ethical principles, assessing societal impact thoughtfully and centering people in the development process, we can reap AI’s benefits while steering clear of its risks.
I am thrilled to present you with a carefully curated collection of over 125 invaluable resources for product managers, combining our best newsletter editions along with other top-notch resources from around the web. This edition aims to be your one-stop shop for insights, frameworks, tools, and techniques to level up your product management game.
As a product manager, one of your biggest responsibilities is understanding your users and iterating your product to meet their needs. However, taking this too far can lead your product into something called the product death cycle.
In this comprehensive guide, we’ll explain what the product death cycle is, why it happens, and most importantly - how you can avoid it as a PM to create a product with happy users that continues to grow.
What is the Product Death Cycle?
The product death cycle refers to the situation where a product development team focuses too heavily on customer feedback rather than having a strong product vision.
This leads to a vicious cycle where:
The product launches but no one uses it initially
The team asks customers what features are missing
They build those features based on feedback
Still, no one uses the product much
So they ask for more feedback and build more features
...And the cycle continues until time or money runs out, leading to the “death” of the product because it never gained traction.
This often happens because the team thinks they are doing the right thing - listening to customers and giving them what they ask for. But good intentions here can lead to disaster.
Share
Why Does This Happen?
There are a few key reasons why PMs and product teams get stuck in this death cycle:
1. Lack of Strong Product Vision
Having a clear, focused product vision is key. As Jeff Bezos says, you need to be stubborn on the vision but flexible on the details.
Without that North Star, it’s easy to get blown off course by every gust of customer feedback. Then you end up with a muddled product trying to be everything to everyone vs. a focused product that delights your core users
.
2. Assuming More Features = More Users
It's a common mistake to think that simply adding more and more features requested by customers will suddenly get tons of people to use your product.
The root problem could be bad positioning, poor messaging, a confusing onboarding flow, bad marketing, poor pricing, or many other issues. Features alone won’t solve those problems.
3. Customers Can't Design Solutions
Customers are great at identifying problems in your product but they rarely can propose effective solutions on their own. That’s your job as a product manager.
Just asking customers what features they want and building them often leads to a hodgepodge product with no coherent vision.
The Complete Guide to Crafting Effective Product Requirements Documents
As a product manager, one of your most critical responsibilities is defining comprehensive product requirements. However, doing so in a way that clearly aligns stakeholders takes thoughtful planning and communication. This is where creating a detailed product requirements document (PRD) becomes invaluable.
In this comprehensive, we'll dive deep into everything you need to know as a PM to create stellar PRDs that drive product development.
A Brief History of PRDs
PRDs have evolved over time along with product management best practices. In the early 2000s, PRDs were often long, tedious Word documents. But today's PRDs are more streamlined, visual, and collaborative.
As a product manager, getting customer feedback is crucial. But you also can't let customers fully dictate your product roadmap. Finding the right balance is key.
In this guide, we'll cover:
Why listening to customers is important
The different types of customer needs
Common traps when listening to customers
When to listen vs when not to listen
Why Listening to Customers Matters
Listening to customers provides many benefits:
Get feature ideas . Customers can suggest new features or improvements to existing ones. This gives product managers insight into what users find frustrating or lacking.
Understand pain points . Customer feedback reveals struggles and pain points when using your product. This shows opportunities to improve the user experience.
Gain market insights . Feedback provides insights into customer demographics, needs, behaviors, and preferences. This intelligence fuels product and marketing decisions.
Clearly, listening to customers provides value. You gain critical insights you'd miss otherwise. But you can't listen blindly.
Share
Understanding the Different Types of Customer Needs
As a product manager, fully grasping your customers' needs is essential for building products that delight you. However, customer needs are complex and multifaceted. They cannot be boiled down to a simple request list.
To get a comprehensive understanding, it's important to segment customer needs into different categories based on how overtly they are expressed. This allows you to dig deeper into the customer psyche to uncover insights across the spectrum of conscious to unconscious desires.
Let's explore four key categories of customer needs: expressed , unexpressed , latent , and future needs . Each offers valuable signals to guide your product strategy if you know where to look.
Expressed Needs : The Tip of the Iceberg
Expressed needs represent the requests and feedback customers directly communicate about your product. This includes:
Feature requests and bug reports
Responses to surveys and interviews
Ratings, reviews, and social media posts
Support tickets and live chat queries
In short, expressed needs encompass any need a customer voluntarily tells you about.
Your AI toolkit breakdown after $4,000 of testing
Claude 3.5 Sonnet → Deep analysis and structured thinking powerhouse Gemini 2.5 Flash → Lightning fast and crazy cheap for high-volume work
Grok 3.0 → Real-time info and uncensored brainstorming Perplexity Pro → Research with proper citations ChatGPT → Your main work partner that remembers everything
My current workflow: • Monday strategy sessions: ChatGPT (remembers our product context) • Quick competitive research: Perplexity Pro • Deep user data analysis: Claude with MCP integration • Real-time market updates: Grok • Technical feasibility: Claude Opus • Document processing: Gemini (handles 750,000+ words)
Think of it this way: ChatGPT is your main work partner. Others are specialized consultants you call for their superpowers
Pro tip: Start with the cheapest model that handles your task. Only upgrade when you hit limitations.
Which model do you rely on most for product work?
Read the complete guide in the comments below